There is increasing interest in using surface plasma waves or other bound local modes to both increase the interaction of incident radiation with detectors and also to provide spectral selectivity. Ongoing work covers both the infrared and the visible spectral regions.
In the visible there is a need for spectral/angular resolution compatible with inexpensive silicon fabrication. While color cameras are ubiquitous, the spectral resolution is provided typically by absorbing dyes, which are increasingly difficult to fabricate with sufficient absorbance as the pixel size decreases, and provide only limited spectral selectivity.
In the infrared, detector signal-to-noise is an important driver. Detectors cooled to cryogenic temperatures provide the highest sensitivity, but require extensive and expensive infrastructure limiting their applicability. Uncooled detectors, typically based on thermal heating of an isolated structure by infrared radiation, have limitations associated with high noise levels and limited response speed.
Surface plasma wave (SPW) and other guided mode interactions provide a method to address some of these issues. There are two related classes of SPWs. For a planar interface between a metal and a device layer (such as a semiconductor material), there is a mode bound to the interface (e.g., a slow wave that is propagating along the interface and evanescent (exponentially decreasing in amplitude) into both the metal and the device layer). This wave is well defined for Re(−Σm)>Re(Σd) where Σm is the permittivity of the metal (with a negative real part) and Σd is the dielectric permittivity. For isolated metal structures, there are localized SPW resonances that concentrate the fields, this localized resonance is involved in the well-known surface enhanced Raman scattering (SERS) effect.
Much of the analysis of SPW modes at a planar metal/dielectric interface has considered only a simple two component structure with a top metal, most often the metal layer incorporates a 1D or 2D grating structure to provide the necessary coupling to the slow SPW and a semi-infinite dielectric layer (e.g., to allow for momentum conservation). However, a realistic detector structure has multiple device layers with varying permittivities. These layers could include contact layers with p- and n-type doping, absorber layers (e.g., quantum dots or strained layer superlattices in the infrared), and electrical isolation layers (often of much lower permittivity than the semiconductor). These layers can have a profound effect on the coupling to, and even to the existence of, the SPW and need to be considered in a full device analysis and design. Additionally, this simple two material model typically does not allow for separate indentification of the useful absorption in the absorber layer and of the parasitic absorption in the metal layer and are hence not sufficient for detailed designs.